Real-time Compressing Algorithm based on Outer-trajectory Measurement Data

• Published in: MATEC Web of Conferences Vol. 63; p. 5034
• Main Authors: Wu, Jin Mei, Ling, Xiao Dong, Hu, Shang Cheng, Wang, Zhen Ping
• Format: Journal Article Conference Proceeding
• Language: English
• Published: Les Ulis EDP Sciences 01.01.2016
• Subjects: Algorithms; Compressing; Compression ratio; Data compression; Data processing; Datasets; Error analysis; Mathematical analysis; Noise filtering; Outer-trajectory measurement data; Polynomial fitting; Polynomials; Random errors; Real time; Trajectory measurement; Truncation errors
• ISSN: 2261-236X; 2274-7214; 2261-236X
• DOI: 10.1051/matecconf/20166305034

Since huge sample datum has to be compressed properly in pre-processing to be sent out, a good compression algorithm will evidently improve the precision of the data-processing. In this paper, a compressing algorithm was studied based on polynomial fitting method. During the process of the real-time trajectory data compression, datasets were successively accumulated according to compression ratio. To apply all the information in the dataset, a series of orthogonal polynomial basis were applied to fitting the function, the least square estimation method was used to filter noise, and the estimated values of the position and the speed from differentiation of object datum in the dataset were sent out as compressed datum. And to get the best filter parameters, the mathematical expression of the error expectations and variances were studied. The compressing principle was given by considering the truncation error and random error simultaneously, which showed that, the best filter was the one by 21-point 3-order polynomial for position data compressing, while for speed data the filter by 41-point 2-order polynomial was better. The theoretical analysis and the simulation results were also provided to prove the effectiveness of this algorithm in data-compression and noise filtering.